[Gluster-devel] GlusterFS High Availability Storage with GlusterFS on Centos5

[Guido Smit]

Hi all,

I've just set up my new cluster with glusterfs. I've tried the examples 
on the site on GlusterFS High Availability Storage with GlusterFS, but 
it keeps crashing, so I've moved several things to the client
config and it's running smooth now. My question now is, what can I do to 
increase performance and are there any mistakes I made in my configs?

My servers all run Centos5 with glusterfs-1.3.pre7 and fuse-2.7.0 (from 
sourceforge).

My configs:

glusterfs-server.vol :

volume mailspool-ds
       type storage/posix                              # POSIX FS 
translator
       option directory /home/export/mailspool         # Export this 
directory
end-volume

volume mailspool-ns
       type storage/posix                              # POSIX FS 
translator
       option directory /home/export/mailspool-ns      # Export this 
directory
end-volume

volume mailspool-ds-io              #iothreads can give performance a boost
       type performance/io-threads
       option thread-count 8
       option cache-size 64MB
       subvolumes mailspool-ds
end-volume

volume mailspool-ns-io
       type performance/io-threads
       option thread-count 8
       option cache-size 64MB
       subvolumes mailspool-ns
end-volume

volume server
       type protocol/server
       option transport-type tcp/server
       subvolumes mailspool-ds-io mailspool-ns-io
       option auth.ip.mailspool-ds-io.allow 192.168.1.*,127.0.0.1
       option auth.ip.mailspool-ns-io.allow 192.168.1.*,127.0.0.1
end-volume


glusterfs-client.vol :

##############################################
###  GlusterFS Client Volume Specification  ##
##############################################
volume pop2-ds
       type protocol/client
       option transport-type tcp/client        # for TCP/IP transport
       option remote-host 192.168.1.42         # DNS Round Robin 
pointing towards all 3 GlusterFS servers
       option remote-subvolume mailspool-ds-io # name of the remote volume
end-volume

volume pop2-ns
       type protocol/client
       option transport-type tcp/client        # for TCP/IP transport
       option remote-host 192.168.1.42         # DNS Round Robin 
pointing towards all 3 GlusterFS servers
       option remote-subvolume mailspool-ns-io # name of the remote volume
end-volume

volume smtp1-ds
       type protocol/client
       option transport-type tcp/client
       option remote-host 192.168.1.224
       option remote-subvolume mailspool-ds-io
end-volume

volume smtp1-ns
       type protocol/client
       option transport-type tcp/client
       option remote-host 192.168.1.224
       option remote-subvolume mailspool-ns-io
end-volume

volume mailspool-afr
       type cluster/afr
       subvolumes pop2-ds smtp1-ds
       option replicate *:2
end-volume

volume mailspool-ns-afr
       type cluster/afr
       subvolumes pop2-ns smtp1-ns
       option replicate *:2
end-volume

volume bricks
       type cluster/unify
       subvolumes mailspool-afr
       option namespace mailspool-ns-afr
       option scheduler alu
       option alu.limits.min-free-disk  60GB              # Stop 
creating files when free-space lt 60GB
       option alu.limits.max-open-files 10000
       option alu.order 
disk-usage:read-usage:write-usage:open-files-usage:disk-speed-usage
       option alu.disk-usage.entry-threshold 2GB          # Units in 
KB, MB and GB are allowed
       option alu.disk-usage.exit-threshold  60MB         # Units in 
KB, MB and GB are allowed
       option alu.open-files-usage.entry-threshold 1024
       option alu.open-files-usage.exit-threshold 32
       option alu.stat-refresh.interval 10sec
end-volume

### Add writeback feature
volume writeback
       type performance/write-behind
       option aggregate-size 131072 # unit in bytes
       subvolumes bricks
end-volume

### Add readahead feature
volume readahead
       type performance/read-ahead
       option page-size 65536     # unit in bytes
       option page-count 16       # cache per file  = (page-count x 
page-size)
       subvolumes writeback
end-volume

--
Regards,

Guido Smit